Brain images application and supervised learning algorithms: a review

Baher H. Nayef, Siti Norul Huda Sheikh Abdullah, Rizuana Iqbal Hussain, Shahnorbanun Sahran, Abdullah H. Almasri

Research output: Contribution to journalArticlepeer-review

Abstract

Medical image processing and classification are important in medicine. Many Magnetic Resonance Images (MRI) are taken for an individual. To reduce the radiologist workload and to enable more efficiency in brain tumor detection and classification. Many Computer Aided Diagnose (CAD) systems have been developed using different segmentation methods and classification algorithms. This study synthesizes and discusses some studies and their results. A Learning Vector Quantization (LVQ) classifier is used to classify MRI images into normal and abnormal. An initial experiment consisting of normal and abnormal MRI Brain Tumor dataset from UKM Medical Center, to observe various versions of LVQ classifiers performance is conducted.From the extensive and informative studies and numerical experiments, it is expected to obtain better brain tumor classification in the future using Multi pass LVQ classifier which obtained the least standard deviation value (0.4) and the mean accuracy rate is equal to 91%.
Original languageEnglish
Pages (from-to)108-122
Number of pages15
JournalJournal of Medical Sciences
Volume14
Issue number3
DOIs
Publication statusPublished - 2014

Fingerprint

Dive into the research topics of 'Brain images application and supervised learning algorithms: a review'. Together they form a unique fingerprint.

Cite this